Dr.
Konstantin Matveev and Ralph Duncan
Development of the Tool for Predicting Hydrofoil
System Performance and Simulating Motion of
Hydrofoil-Assisted Boats
ABSTRACT
Hydrofoil technology is experiencing re-
emergence due to demands for higher performance
in both military and commercial sectors.
Hydrofoil-assisted ships and boats use foils to
pa rt
iallyorf ul
lys uppo r tas hipswe ight .To
reduce hydrodynamic resistance, a significant
fraction of the ship hull is lifted out of the water.
Hydrofoils can also be very effective in mitigating
motions in rough seas. One of the factors limiting
wider application of the hydrofoil technology is
difficulty and high cost of developing high-
performance hydrofoil systems. A set of computer
programs that will improve design of hydrofoil-
assisted boats is currently under development.
This tool can be used for initial optimizing of
complex hydrofoil configurations and for
developing hydrofoil control systems. Steady and
unsteady forces on hydrofoil system elements are
represented in the general form suggested by
fundamental hydrodynamic theories. Empirical
corrections account for viscosity and other effects.
Interactions between hydrofoil systems are
included. The motion simulator can predict
stability and maneuverability of hydrofoil-assisted
boats and their motions in waves.
INTRODUCTION
Hydrofoils are characterized by the highest lift-
drag ratio among all types of water-borne craft
within the optimal regime for hydrofoils.
Hydrofoil technology attracted significant
attention in the second half of the twentieth
century and thousands of hydrofoil boats were
built. However, hydrofoil ships were displaced
from a dominant position on the fast ferry market
by simpler and lower-maintenance catamarans that
are more efficient at larger relative dimensions (or
lower relative speeds).
The hydrofoil technology is experiencing re-
emergence. Retrofitting existing catamarans with
hydrofoils and building new hydrofoil-assisted
multi-hulls is a booming industry. Hydrofoils have
significantly improved the efficiency of
commercial fast ferries. Hydrofoil systems are
successfully applied for motion mitigation on large
fast ships and show similar promise for future
Sealift ships. Although there are currently no
serious attempts to introduce hydrofoil technology
to the pleasure boat market, hydrofoils have
reduced fuel consumption on small boats by nearly
half and considerably improved seaworthiness.
One of the problems preventing a wider use of
hydrofoils is the complicated development process
of efficient, robust, and inexpensive hydrofoil
systems.
There are many publications on specifics of
hydrofoil modeling. Some of the previous studies
used a simplified approach (e.g., Sakic 1981,
Latorre and Teerasin 1992) that did not account
for many important factors, such as effects of
motion history and interactions between foil
systems. On the other hand, Computational Fluid
Dynamics tools have recently been developed that
can accurately simulate hydrofoil systems (e.g.,
Walree 1999, Migeotte 2002). However, these
tools are computationally intensive and still
partially rely on empirical knowledge.
This paper describes efforts to develop relatively
simple engineering tools for practical
hydrodynamic design of hydrofoil systems and
boats. The basic idea is to represent the
hydrodynamic forces in the form suggested by
fundamental theories. Correlation factors are
introduced that account for real-life effects
neglected by theories. These factors are obtained
either from other mathematical models addressing
specific phenomena or from empirical knowledge
(e.g., Ogilvie 1958, Egorov and Sokolov 1965).
An essential step in developing these engineering
tools is validating them against experimental
results. One of the requirements in this project is
to keep the development cost low and the tool
simple enough for use in practical parametric
design. Therefore, some discrepancies between
model results and test data can be tolerated.
Design of hydrofoils and other advance marine
vehicles is not as straightforward as that of
conventional (usually slow) ships. Model testing is
strongly recommended. Intermediate-scale
prototypes are also useful for reducing the
uncertainty of scale effects and in achieving
optimal performance at the full scale. The
engineering tool being developed is not considered
as a substitute for testing, but as a means for
preliminary optimizing of the ship design and
reducing development time.
Physical phenomena most critical for realistic
design of high-performance hydrofoil systems are
outlined in the following sections. In particularly,
a general formula for the lift of a hydrofoil section
is given and its generalization to complex
hydrofoil systems is outlined. We also discuss a
tandem foil interaction, a convenient method for
calculating unsteady hydrodynamic forces, and a
development of 6D motion simulator for a
hydrofoil boat. Correlations and functions are not
presented here in details due to paper size limits
and because of the proprietary nature of some
correlations and methods.
STEADY HYDRODYNAMIC
FORCES
The lift force generated on a hydrofoil (Figure 1) is
determined by formula
U 2
Z C z S, (1)
2
where C z is the lift coefficient, is the water
density, U is the flow velocity with respect to the
foil, and S is the one-side foil area. S b 2 ,
where is the foil aspect ratio and b is the
effective chord. The lift coefficient of a hydrofoil
with finite aspect ratio, moving under free water
surface is
Z
z
h
X
x
b U
Figure 1: General scheme of a submerged hydrofoil.
C
k z 0 0
Cz 0
. (2)
C z k 1
1
0
The derivative of the lift coefficient on the attack
Cz
angle is for a deeply submerged, infinite
0
aspect ratio hydrofoil. For the ideal fluid and thin
foil profiles this coefficient is equal to 2. A
correction to this coefficient that depends on the
profile thickness-chord ratio, Reynolds number,
and the trailing edge closure angle should be
applied to account for these effects (e.g., Martin
1963). The apparent attack angle is the angle
between the line connecting trailing and leading
edges and the horizontal plane. The effective zero-
attack angle is 0 . This angle is a function of the
foil camber, Reynolds number, and the trailing
edge closure angle (Egorov and Sokolov 1965).
The XFOIL program (Drela 1989) is another
method for determining the effect of viscosity on
the zero-attack angle. Discussions on the effect of
Reynolds number can be found in (Walree 1999)
and (Migeotte 2002). The proximity to the free
water surface is accounted for by corrections 0
and k to the effective attack angle and to the lift
derivative, respectively. They are functions of the
profile thickness-chord and submergence-chord
ratios (Egorov and Sokolov 1965). Denominator in
Eq. (2) is due to a finite aspect ratio ; and is
Glauert correction. The influence of the foil
submergence that affects the vorticity wake is
accounted by and the influence of struts is
accounted by (Egorov and Sokolov 1965). This
is a function of the strut positions and the foil
submergence and aspect ratio.
Figure 2 compares the results of Eq. (2) and
experimental data for the lift coefficient of a
hydrofoil section at Reynolds number 2.5x106.
This figure illustrates the uncertainties one should
expect when using approximate mathematical
models.
employed on modern hydrofoil fast ferries (Figure
3), we propose to apply the transverse strip
approach. The foil span is divided into a number
of sections and the local lift coefficient is
computed for each of the sections considering the
local foil profile, local flow characteristics, and the
global span of the hydrofoil system.

0.5

0.4
Lift coefficient

0.3

0.2

0.1

-0.1
-2 0 2 4
Installed attack angle [deg]

Figure 2: Lift coefficient for profile YS-920. ,

experimental data for smooth and rough surface
(Shen 1985); solid line, model prediction.

Figure 3: Photo and front view of a hydrofoil fast

Hydrodynamic resistance of a hydrofoil system
includes profile, inductive, and wave drag
components. Struts have profile, spray, and
sometimes cavitation drag components (Egorov
and Sokolov 1965, Voitkunsky 1985). Hydrofoil
sections crossing the water surface may also
generate spray drag. Special fences are usually
installed on hydrofoil and strut sections near the
operational waterline to minimize this drag
component and to avoid foil ventilation. High-
performance hydrofoils should operate in the
subcavitating regime. The boundary of the
subcavitating domain for a particular profile is a
function of the lift coefficient, profile thickness-
chord ratio, and cavitation number.
For calculating hydrodynamic forces on the
complex surface-piercing foil systems, such as
ferry of Olympia class.
One of the initial variants of the bow hydrofoil
system of Olympia-class ship (Figure 3) was tested
on a model scale at chord-length Reynolds number
5.5-8.0x105. Measured lift and drag are presented
in Figure 4 by open circles. Predictions for this
system by the theory outlined in this paper are
shown by crosses. (Some details of the foil system
geometry are known to us only approximately.)
Adequate agreement between model and test
results is observed in Figure 4 in the range of
operational attack angles and submergences.
Underprediction of the lift can be partly attributed
to a reduced viscous influence on the lift
derivative and a zero-attack angle in the vicinity of
free water surface (Migeotte 2002). Another
important factor is the non-uniform distribution of
the lift force along the span of a hydrofoil. On a
single zero-dihedral hydrofoil, the central sections
are more heavily loaded than the sections near the
tips. Therefore, a hydrofoil with deeper
submergence of the central part will generally
produce higher lift than predicted by the approach
outlined here, although the influence of struts will
partly compensate for this effect. Simple semi-
empirical corrections can be applied to account for
such phenomena in engineering design of
hydrofoil system.
1.5
+ 4 deg
1
Lift [KN]
upwash region in the wave system, then the lift
force generated on this foil will incline forward
effectively creating thrust. The length of the wave
hollow is approximately proportional to the speed
and a square root of the span of the front foil. The
deviation of the water surface from the
undisturbed level depends on the front foil
geometry, its submergence, lift coefficient, and the
ship speed. Expressions for engineering
calculations of these values are available
(Kolyzaev, Kosorukov, and Litvinenko 1980, Bai-
Qi 1981, Voitkunsky 1985). Besides the effect of
deforming the water surface, there is also the
influence on the stern foil by the trailing vortices
of the front foil. For shallow submerged
hydrofoils, the influence of the free water surface
should be accounted for when estimating this
effect (e.g., Voitkunsky 1985).

0.5 0 deg

- 4 deg
0

1 1.5 2
Relative submergence

0.1

0.08 Figure 5: Wave hollow behind a hydrofoil and

+ 4 deg tandem hydrofoil arrangement.
Drag [KN]

0.06

0.04 Tandem hydrofoil systems have found widespread

0 deg applications in the second half of the last century
0.02
(Matveev K.I. and Matveev I.I. 2001). However,
0 the complexity of their development and
1 1.5 2 production is a significant barrier to introducing
Relative submergence
new ships. Special maintenance is also required
(Matveev 2001). Hydrofoil-assisted ships, where
Figure 4: Lift and drag generated on the model-scale the weight is only partially supported by
fore foil system of one experimental variant of hydrofoils, are becoming more popular. Such ships
Olympia class hydrofoil. experimental data use simpler hydrofoil systems, and a portion of the
(Matveev I.I. 1999); theoretical results. hull always remains in contact with water,
simplifying the propulsion system arrangement.
Interactions between hulls and foils on such boats
FOIL INTERACTIONS are complicated, and optimization of the
performance of new configurations requires
A hydrofoil moving in the proximity of the free complex analysis or elaborate testing programs
water surface creates a wave system behind it (Migeotte 2002).
(Figure 5). If another hydrofoil is placed in the
One high-performance tandem-type configuration
is shown in Figure 6. It is especially attractive for
shallow-water or heavy ships. The front foil
creates a wave hollow behind itself, and a planing
stern with a propulsion system is located in the
favorable upwash flow region. The middle foil
system consists of two separate foils at the sides of
the boat. These mid foils operate outside the wave
hollow generated by the front foil. The location of
the mid foils can be chosen to provide favorable
front-to-middle foil interaction. Mid foils can also
augment the upwash flow at the planing surface.
Therefore, all interactions between hydrodynamic
elements in this configuration are favorable. Other
significant advantages of this system include
improved roll stability, more uniform distribution
of hydrodynamic support along the ship structure
(which reduces bending moments in the hull), and
better seakeeping.
Figure 6: Scheme of a boat with positive interactions
between hydrodynamic elements (Maksimov et al.
1975).
UNSTEADY FORCES
When predicting unsteady motions of a boat, one
must know unsteady hydrodynamic forces on
hydrofoils. If unsteady processes are sufficiently
slow (e.g., in long waves), then a quasi-steady
approach can be applied to calculate forces using
Eqs. (1-2). If disturbances are small and periodic,
then the forces (both amplitude and phase) will
depend on Strouhal numbers (or frequency). In
essentially unsteady problems, the fully unsteady
methods for calculating forces should be applied.
Vortex-lattice methods have been used for
hydrofoils (e.g., Walree 1999), but the complexity
of this technique makes it inconvenient for
parametric design studies of multi-component
hydrofoil systems. In our approximate approach,
we use analytical results obtained for simple
problems with added empirical corrections
accounting for real-life effects. For example, a
general form of the unsteady hydrodynamic lift on
a hydrofoil can be expressed as follows (Egorov
and Sokolov 1965)
Y Yqs Yam Yvw . (3)
The quasi-steady force Yqs is calculated by Eq. (1)
but with the effective attack angle
ef V1 b/ 4 / U , where is the angular
velocity along the transverse axis, V1 V y U,
V y is the vertical velocity of a hydrofoil, and is
the apparent attack angle.
d
mV1
Yam . (4)
dt
This is the force of an inertial nature that depends
on the added mass. For rectangular hydrofoils, the
added mass can be approximated as for a plate
with corrections for the finite aspect ratio and
relative submergence
m
b / 2 bk () f1 (h / b) .
2
The last force depends on the shed vorticity
1
Ysv b 2 U f 1 (h / b)
2
( s)ds
s1
, (5)
0 s s1 b / 2
2
b / 2
2
where (s ) is the circulation density in the wake
at a horizontal coordinate s , and s1 is the current
coordinate of the hydrofoil trailing edge. For
several classical cases (e.g., harmonic variations,
steady acceleration, etc.) and idealized hydrofoil
profile, the intensity of shed vorticity can be
determined analytically; in other cases it can be
found numerically in a time stepping process. This
method for calculation of unsteady hydrodynamic
lift gives results in acceptable agreement with
experimental data (Egorov and Sokolov 1965).
BOAT DYNAMICS
The ultimate goal of the model under development
is motion prediction of a hydrofoil craft in both
calm water and in unsteady operations, including
rough seas. The motion simulator integrates the
boat dynamics equations applying the expressions
for the forces outlined above. When conducting
such modeling, two or more coordinate systems
are usually used. For example, equations of
motions are written in the ship-fixed coordinate
system where inertial moments do not change.
Motion trajectory is presented in the space-fixed
system. When disturbances are small, the
equations can be simplified. Sometimes the
problem can be reduced to motions in the
coordinate sub-spaces, such as the vertical plane
motion (pitch, heave, surge) for seakeeping in
head and following waves, or lateral motion
(sway, yaw, roll) for maneuvering. Our tool
incorporates all degrees of freedoms, with the
ability to study reduced-order problems.
The general form of the dynamics equations are
written as
M M
x results that may be needed for design optimization
F ( x, x
; history; waves; control ) , (6)
where M is the inertial matrix, M is the added
mass matrix, and x is the vector of six
coordinates. The generalized force vector on the
right-hand side of Eq. (6) depends on the position
and velocities of the craft, history of its motion,
sea waves, and control system actions (e.g., flaps
on hydrofoils). Besides forces acting on the
hydrofoils, the forces generated on the appendages
and propulsors, as well as the forces due to above-
water hull motion in still or windy air, are
included.
An interesting phenomenon in the dynamics of a
hydrofoil craft is the unsteady fore-aft foil
interaction. In complete CFD methods, these
interactions are accounted for by the vorticity
shedding from and the waves produced by the
forward foil. In our simplified approach, only the
unsteady interaction due to the water surface
deformation is implemented, similar to (Kaplan
1955). This mechanism is dominant for surface-
piercing and shallow submerged foils. A variation
in the water surface deformation produced by the
forward foil is felt by the aft foil after time delay
L / U , where L is the distance between
hydrofoils. This approximation is sufficient for
useable engineering results.
The dynamic system outlined here is applied for
simulating vertical plane motion in waves of the
model of one variant of Olympia-class hydrofoil.
Experimentally obtained steady lift and drag
coefficients (which are functions of attack angle
and submergence) were used with the addition of
unsteady forces, wave-induced forces, and the
fore-aft foil interaction. Calculated amplitudes of
heave and pitch in following waves are compared
in Table 1 with experimental data and with results
obtained by simpler (but still nonlinear) theory
which neglects vorticity wake and the fore-aft foil
interaction. Following waves are usually the most
adverse wave direction for a hydrofoil craft. It
should be noted that the final, optimized
configuration of the Olympia hydrofoil, which
additionally employs controlled flaps, has motions
in high seas 2-10 times lower than those for the
hydrofoil system variant studied here. Higher
accuracy between test data and experimental
will require using sophisticated CFD methods for
modeling hydrofoil craft dynamics.
Motion parameter Heave Pitch
Experimental data 35 mm 2.1 deg
(Matveev I.I. 1999)
Prediction by the 30 mm 2.4 deg
presented theory
Prediction by the 38 mm 2.8 deg
theory neglecting
foil interaction and
vorticity wake
Table 1: Amplitudes of heave and pitch in following
waves of the model of one experimental variant of
Olympia class hydrofoil. Speed 5.4 m/s (Fr D = 2.7);
wave height 130 mm; wave length 3.25 m.
CONCLUDING REMARKS Ma r
tin,M. ,Thes t
ability derivatives of a
hydrofoilboat,
Hy drona uti
csI nc orporate d,
An engineering tool for calculating hydrodynamic Technical Report 001-10, 1963.
forces on hydrofoil systems and for modeling boat
dynamics is being developed. It will be used for Matveev, I.I. (chief designer of Olympia class
designing hydrofoil-assisted craft. This relatively hydrofoils), personal communications, 1999.
simple approach is suitable for parametric studies Ma tveev,K.I.
, Mainte
nan c eofhy drofoil
of the influence of foil elements on hydrodynamic systems,
2ndI nt e
rnati
ona l Conference on High-
performance, seakeeping, and maneuverability. Performance Marine Vehicles, Hamburg,
Controlled foil sections that improve boat Germany. 2001.
performance will also be incorporated into the
tool. Planing and semi-planing hull elements will Ma t
v e
e v,K.I.a ndMatveev,I
.I.
,Ta nd em
be added to model transitional regimes of pure hydrofoi
ls yst
e m,Ocean Engineering, 28(2), pp.
hydrofoil boats and service regimes of hydrofoil- 253-261, 2001.
assisted ships. Further development of the tool Migeotte, G., Design and Optimization of
towards CFD will also be considered. Hydrofoil-Assisted Catamarans, PhD Thesis,
University of Stellenbosch, South Africa, 2002.
Ogilvie,T.
F., Thet heoreticalpred i
cti
onofthe
REFERENCES l
ong i
tudinalmot ionsofhy d rofoi
lc raft
,Dav
i d
Bai-Qi,L.,Ap r
e dict
ionme t
hodoff oilbo rne Taylor Model basin, Report 1138, 1958.
performance characteristics of hydrofoil craft in Sakic,V. ,Approxi ma tede t
erminati
onofthe
calms ea,
Hi g hSpe edSu r
fa ceCr aftConf erenc e, propulsive power of sma l
lhy drof
oilcra
ft,
High
London, 1981. Speed Surface Craft, 1982.
Drela
,M. , XFOI L:Ana na l
y si
sa nddesi
g n
Shen,Y.T. ,Wi ngs ectionsf orhy dr
ofoispart 3:
l
sy
stemf orlowRe ynoldsn umbe ra i
rfoi
ls,

experi
me nt a
lv erificati
ons ,
Journal of Ship
Conference on Low Reynolds Number Airfoil
Research, 29(1), pp. 39-50, 1985.
Aerodynamics, University of Notre Dame, 1989.
Voitkunsky, Y.I., Handbook on Ship Theory,
Egorov, I.T. and Sokolov, V.T., Hydrodynamics of
Sudostroenie, Leningrad, 1985. (in Russian)
Fast Craft, Sudostroenie, Leningrad, 1965. (in
Russian) Walree, F. van, Computational Methods for
Hydrofoil Craft in Steady and Unsteady Flow,
Ka p lan,P., Thefo r
cesa ndmome nt sactingona
PhD Thesis, Delft University of Technology,
tande mhy dr of
oilsystemi nwa v es
,St eve ns
Netherlands, 1999.
Institute of Technology, Report No. 506, 1955.
_______________________________________
Kolyzaev, B.A., Kosorukov, A.I., and Litvinenko,
V.A., Handbook on Design of Dynamically
Supported Ships, Sudostroenie, Leningrad, 1980. Dr. Konstantin Matveev o btainedh i
sMas t
er
s
(in Russian) degree in Applied Physics from Moscow Institute
of Physics and Technology and his Ph.D. degree
La torre ,R. a ndTe eras i
n,S. ,Cal
culati
ono f in Mechanical Engineering from California
hydrofoil craft take-off speed including influence Institute of Technology. He has 15 years of
of foil size, foil angle, and propel
lershaftsize, research and consulting experience in
Ocean Engineering, 19(2), pp. 183-197, 1992. hydrodynamics, acoustics, and thermal sciences.
Maksimov, V.Y., Matveev, I.I., Volkov, V.V., He worked on hydrodynamics and dynamics of
Korotkov, M.M., Firsel, V.B., and Erlykin, I.I., advanced marine vehicles: air-cavity ships,
Hi gh-s pe edc r aft,
I nnov ationCerti
fi
ca t
e4 72056, planing craft, wing-in-grounds (WIG), ships with
USSR, 1975. (in Russian) aerodynamic unloading, multi-hulls, SWATH, and
hydrofoils. As a post-doctorate, he advanced
thermoacoustic technology at Los Alamos
National Laboratory. Currently Dr. Matveev is
Senior Hydrodynamicist at Art Anderson
Associates. He leads research efforts on high-
performance marine vehicles and conducts
fundamental research in the area of fluid
mechanics.
Mr. Ralph Duncan is a retired US Navy
Engineering Duty Officer with more than thirty
years of maritime experience. His broad-based
experience provides a unique blend of ship
handling, water transportation, construction
management, marine engineering and design, and
marine project management experience as well as
formal graduate training in design of ship systems.
In the last ten years he has applied his knowledge
and experience to marine projects that have
included a foil-assisted passenger-only ferry
design, a 52 foot high speed aluminum passenger
vessel, investigation of ferry wake wash, ferry
system analyses, and marine engineering services
for a large vehicle ferry. At Art Anderson
Associates he has held positions as Senior Marine
Engineer, Vice President of Business Development
and Vice President of Marine Engineering. He is
currently Principal for Military and Innovative
Vessel Design and Program Manager for the
company Seabase and Sealift programs.